Collapse by mean and mode

[Dakota McAvoy]

Hello, I am currently working with a dataset that has both continuous and dichotomous or categorical variables. I am hoping to collapse them by individual within series, but as far as I am aware, -collapse- does not take mode. Is that correct?

So with the data below, if I did:

collapse(mean) var1 var2 var3 var4, by(id)

It would work wellfor var2, which is continuous, but not for var1/var3 (which are dichotomous) or var4 (which is categorical). For those, I would ideally want to take the mode, not the mean. Does anyone have a good workaround for this?

* Example generated by -dataex-. For more info, type help dataex
clear
input byte(var1 var2 var3 var4 id series)
1 56 1 1 1 1
1 45 1 2 2 1
1 65 0 2 3 1
1 88 1 4 4 1
0 89 0 3 5 1
1 52 0 1 1 2
1 46 1 2 2 2
1 65 0 3 3 2
1 88 1 4 4 2
0 89 0 3 5 2
1 53 0 1 1 3
1 45 1 2 2 3
1 65 0 2 3 3
0 88 1 4 4 3
0 89 0 3 5 3
end

[William Lisowski]

See

help egen

which has a mode() function.